Test Bed Cases Attempting to Duplicate Test Results

December 12, 2014

Summary

• Enhanced cooling at the 12 mm argon layer can’t explain more than an 0.5°K drop in temperature but the heater layer drops 0.75°K– The aluminum plate side of the argon layer is +0.54°K relative to the

bulk argon so it can only drop by this amount and nothing can drop by more than this as a result.

• Increased cooling of the 3 mm copper plate could result in large, rapid temperature drops at the heaters.– Nucleate boiling has high enough convection– The test result indicates some sort of triggering event that results in

increased heat removal.• Superheated argon and nucleate boiling?• Argon gas rising up the back of the copper plate after boiling off at the lower edge?

(but I think the back of the copper plate is bonded to the G10 insulating plate).

Heat to State Just Before Sudden Change Under Test

1

90

90.2

90.4

90.6

90.8

91

91.2

91.4

91.6

91.8

92

VALU

0400

8001200

16002000

24002800

32003600

4000

TIME

heater

Al_plate

Cu_matrix

Test bed to measure convection in Argon

Continuation of test with unchanged boundary conditions1

90.2

90.4

90.6

90.8

91

91.2

91.4

91.6

91.8

92

92.2

VALU

0400

8001200

16002000

24002800

32003600

4000

TIME

heater

Al_plate

Cu_matrix

Transition from laminar to turbulent (2x H)

Temperatures continue to rise slowly, they are asymptotically approaching steady state.

Continuation of test, Argon layer transitions to turbulent heat transfer1

90.25

90.375

90.5

90.625

90.75

90.875

91

91.125

91.25

91.375

91.5

VALU

0400

8001200

16002000

24002800

32003600

4000

TIME

insul_1

insul_2gap_1

gap_2

Cu_2

Transition from laminar to turbulent (2x H)The heater layer drops ~0.06°K in 0.17 hours In the test the heater layer dropped 0.8°K in ~0.01 hours

The conduction across the argon gap is doubled for turbulent heat transfer.

Continuation of test, Nucleate Boiling on the aluminum face at the Argon layer

The heater layer drops 0.125°K in 0.17 hours In the test the heater layer dropped 0.8°K in ~0.01 hours

The conduction across the argon gap is increased ten fold to represent nucleate boiling.

The surface of the aluminum plate cools ~0.23°K in ~0.02 hours.

Even if the argon gap conduction was infinite the heater layer temperature would not drop much more.

1

90.25

90.375

90.5

90.625

90.75

90.875

91

91.125

91.25

91.375

91.5

VALU

0400

8001200

16002000

24002800

32003600

4000

TIME

heater

Al_plate

Cu_matrix

Transition from laminar to turbulent (2x H)

Continuation of test, Nucleate Boiling on the vertical edges of the copper plate

The heater layer drops 0.9°K in ~0.1 hours In the test the heater layer dropped 0.8°K in ~0.01 hoursThe heaters were turned off at 0.13 hours in this run.

The vertical edge of the 3mm copper plate is given a convection coefficient of 10,000 W/m2-°K (see next slide)

The copper plate seems to be cooling more slowly than the test.

Is this plate exposed to Argon on its backside?

1

90

90.2

90.4

90.6

90.8

91

91.2

91.4

91.6

91.8

92

VALU

0400

8001200

16002000

24002800

32003600

4000

TIME

heaterAl_plateCu_matrix

Nucleate boiling on OD

Nucleate Boiling

The smaller values of Csf (0.003) give h=10,000 W/m2-°K for a 1°K differential and could explain the test results if applied to the vertical edges of the copper plate.

Note that the sudden offset and subsequent cooling seem to require that the Argon be superheated and for nucleate boiling to continue once triggered.

Nucleate Boiling Calculation

Csf = 0.003 to 0.015 for listed surface and liquid material combinationss = 1 for water 1.7 otherwise

pl = 1440 kg/m 3pv = 5.0 kg/m 3 0.31 lb/ft^ 3

g = 9.8 m/sec 2gc * sigma =0.012 gc =1 sigma = 0.012 N/m at 90 K

Kl = 0.1315 W/m/Khfg = 161000 W/sec/kgCp = 1078 j/kg/Kmu = 0.00026 kg/m-sec

Csf = 0.003Tw-Tsat= 0.1 0.2 0.4 0.8 1h = 106 425 1,699 6,796 10,619 W/m 2